The key elements of Figure 1 focus on how Baseline SRE5-LR and impulsivity (IMPULSE) relate to each other and to Time 3 alcohol problems (ALC PROBS). In the model, manifest variables (e.g., age) are rectangles and latent variables (i.e., factors generated within the measurement model) are circles. Based on prior studies, we hypothesized that Time 1 higher SRE5-LR scores (i.e., lower LRs per drink) and higher impulsivity will each relate to higher Time 3 alcohol problems, but will only weakly relate to each other. The model further proposes that SRE5-LR will also relate directly or indirectly to all three Time 2 measures, which in turn will relate directly to ALC PROBS, and indirectly to ALC PROBS through drinking to cope (COPE). In order to evaluate whether SRE5-LR and impulsivity operate similarly within the SEM, relationships similar to those for SRE5-LR are proposed for impulsivity in Figure 1. These key relationships among SRE5-LR, impulsivity, Time 2 variables and ALC PROBS are indicated in bold in Figure 1. Sex, age, ethnicity and AUD histories in one or both parents are used as covariates to determine if the key relationships in bold operate even after considering these additional characteristics.
